Based on the paper Siamese Recurrent Architectures for Learning Sentence Similarity by Jonas Mueller and Aditya Thyagarajan
Based on implementation by fionn-mac
Download pretrained word2vec model from here
Download Quera Dataset here
Trained models with different configurations can be downloaded here
the data and word2vec model should be put into data directory and the pretrained models should be put into output directory
install requirement by
pip3 install -r requirement.txt
run the line below to train
python3 training.py
run the line below to test on validate set
python3 evaluate.py
modify hyperparameters.py to change hyperparameters and select pretrained model